jammorning all05:24
Chipacamorning all!09:02
facubatistaMuy buenos días a todos!10:56
Chipacafacubatista: heyy10:57
facubatistaChipaca, meeting?11:01
Chipacafacubatista: omw11:03
facubatistaChipaca, so, there's no "easy" or "integrated" way to make argparse to support a longer help for the command alone12:29
Chipacafacubatista: i'm listening12:29
facubatistaI mean, if I do `charmcraft version -h`, the machinery of argparse triggers in, not allowing us to have contron on the output (unless we keep digging with hacking its internals), and it doesn't have a field to show a "long description"12:31
facubatistaChipaca, so, we could keep digging, but maybe is to first understand which are all the changes we'll need to do regarding help messages... we may totally craft a HelpFormatter if it's very different than the default12:32
facubatistaChipaca, IOW, what I propose is to forget about charmcraft#3 until we have news and directions from UX12:32
mupIssue charmcraft#3: Support for long help texts from each command <Created by facundobatista> <https://github.com/canonical/charmcraft/issues/3>12:32
Chipacafacubatista: we already have that direction12:33
Chipacafacubatista: but i'm ok with slipping charmcraft#3 due to it turning out to be more work12:33
mupIssue charmcraft#3: Support for long help texts from each command <Created by facundobatista> <https://github.com/canonical/charmcraft/issues/3>12:33
facubatistaChipaca, ok12:36
Chipacai'm going to go figure out lunch before it's tea time12:46
facubatistaChipaca, please remember this: https://github.com/canonical/charmcraft/pull/5713:11
mupPR charmcraft#57: Make pip list current packages before proceeding with the install <Created by facundobatista> <https://github.com/canonical/charmcraft/pull/57>13:11
facubatista(note I re-did the PR, now from the forked project)13:11
mupPR operator#338 opened: ops/storage.py: Move SQLiteBackend <Created by jameinel> <https://github.com/canonical/operator/pull/338>13:11
Chipacafacubatista: how could i forget13:19
Chipaca(answer: easily)13:19
Chipacasounds a bit rude13:19
Chipacapoor http213:20
* Chipaca close to having a working snap13:21
Chipaca9.1MB for a working snap including git and enough helpers to pull https13:30
Chipacafacubatista, jam, standup?13:34
jamChipaca, I'll need your +1 on #338 at some point :)14:14
mupPR #338: ops/storage.py: Move SQLiteBackend <Created by jameinel> <https://github.com/canonical/operator/pull/338>14:14
Chipacajam: you've got it14:22
Chipacajam: facundo'll need your +1 on charmcraft#57 also :-D14:23
mupPR charmcraft#57: Make pip list current packages before proceeding with the install <Created by facundobatista> <https://github.com/canonical/charmcraft/pull/57>14:23
Chipacafacubatista: meeting?14:32
Chipacafacubatista: booooooo you missed the meeting boo14:46
* Chipaca boos14:46
facubatistaChipaca, ugh14:49
facubatistathere's something wrong with my notifications today14:49
* Chipaca boos some more14:49
facubatistadamn, I super wanted to be in that meeting14:50
facubatistaok, pulseaudio is sending the irc audio notification through the headset14:52
facubatistaeven if there's no headset connected :/14:52
* facubatista will need to reboot, probably, to fix this14:54
facubatistaChipaca, any particular outcome/progress?15:01
Chipacafacubatista: the good news is, there is another one next week15:18
Chipacafacubatista: also, we should probably consider moving to click15:19
facubatistaChipaca, ack15:25
Chipacafacubatista: big asterisks on the click thing though :)15:25
Chipacafacubatista: charmcraft#58 might be of interest to you16:56
mupPR charmcraft#58: changes to snapcraft.yaml, and version, for 0.2.0 <Created by chipaca> <https://github.com/canonical/charmcraft/pull/58>16:56
Chipacafacubatista: my devious plan is to land charmcraft#57 to master before landing that16:56
mupPR charmcraft#57: Make pip list current packages before proceeding with the install <Created by facundobatista> <https://github.com/canonical/charmcraft/pull/57>16:56
facubatistaChipaca, I'm *so* glad you're doing the snapcraft part :p16:57
Chipacafacubatista: I am too. At this point I have Opinions about it.16:58
facubatistaChipaca, I wonder how much of that hardcoded listings will change in time; it will be a maintenance problem?16:58
Chipacafacubatista: worst case, we delete the whole stanza and the snap grows 5×16:59
Chipacanot a terrible end16:59
Chipacafacubatista: for now, it works like this16:59
Chipacafacubatista: at some point maybe there'll be a better way16:59
Chipaca(like being able to call out to a git snap)16:59
Chipacaalmost all the really ugly bits is git17:00
Chipacaalthough picking and choosing python libs is also a pain, it's not as random17:00
Chipacafacubatista: for reference, I've done this with the icdiff snap, and it hasn't broken nor needed updated since I created it a couple of years ago17:01
Chipacaneeded updating*17:01
facubatistaChipaca, nice17:01
Chipacafor 0.3, tab completion17:01
Chipacaicdiff, since 2018-09 (so just under 2 years)17:02
Chipacafiled charmcraft#59 for that17:03
mupIssue charmcraft#59: tab completion in-snap <Created by chipaca> <https://github.com/canonical/charmcraft/issues/59>17:03
mupPR operator#338 closed: ops/storage.py: Move SQLiteBackend <Created by jameinel> <Merged by jameinel> <https://github.com/canonical/operator/pull/338>17:35
jamChipaca, facubatista operator#323 updated17:39
mupPR #323: 317 state get <Created by jameinel> <https://github.com/canonical/operator/pull/323>17:39
facubatistaChipaca, I don't like pointing the user to the logfile when CommandError17:51
facubatistaChipaca, or at least we need another exception that is "ended in error", but more gentle17:53
facubatistafor example, it doesn't make sense to point the user to the logs when:    raise CommandError("the requirements file was not found: {!r}".format(str(fpath)))17:53
facubatistabut it does make sense when:           raise CommandError("problems installing dependencies")17:54
facubatistamaybe we need a CommandCrashed (pointing user to the logs) and CommandError (more gentle)17:54
facubatistawhat do you think?17:55
Chipacafacubatista: sounds like it, yes18:46
Chipacafacubatista: the difference being, i think, that in one case we can easily summarise the problem in Human, whereas in the other we can't18:47
facubatistaChipaca, the difference for me, which I'm not finding how to express in the name of the error class, is "here's a problem that I can explain you and give you some hint about how to solve", and "we found a problem because something exploded beneath us and we don't know really what or why"19:04
Chipacafacubatista: KnownError(CommandError)?19:04
facubatistathe only case so far for the second case is when installing dependencies19:05
facubatistaeven a error from the store is a "controlled" one19:05
facubatistammm... IF the store is really telling us an error in a field, it's a controlled one19:06
facubatistaif the store explodes in unpredictable ways, is also the second case19:06
facubatistaCommandControlledError vs CommandUnknownError19:06
facubatistaThisIsSomewhatExpectedError vs WhatWillWeDoWithAllThisBloodError19:07
facubatistaChipaca, so, release?19:58
Chipacafacubatista: looks like it20:45
Chipacafacubatista: think you can write up a release blurb?21:30
Chipacafacubatista: for https://github.com/canonical/charmcraft/releases/edit/untagged-c8c90d68c4d08c8e254721:30
=== ChanServ changed the topic of #smooth-operator to: general discussion of the operator framework || github.com/canonical/operator || ops 0.6.1 || charmcraft 0.2.0
Chipacafacubatista: edit https://github.com/canonical/charmcraft/releases/tag/0.2.0 when you can (or i can do it tomorrow)21:35
facubatistaChipaca, I can do it, but tomorrow21:38
Chipaca0.2.0 in github, pip, and snap21:40
Chipacafacubatista: ⁵21:40
ChipacaEOD here21:43
* facubatista eods too21:45

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!